Synthesis and characterization of manganese oxides employed in VOCs abatement
-
Upload
luciano-lamaita -
Category
Documents
-
view
217 -
download
5
Transcript of Synthesis and characterization of manganese oxides employed in VOCs abatement
Synthesis and characterization of manganese oxides employed
in VOCs abatement
Luciano Lamaita, Miguel A. Peluso, Jorge E. Sambeth *, Horacio J. Thomas
Centro de Investigacion y Desarrollo en Ciencias Aplicadas (CINDECA) ‘‘Dr. Jorge J. Ronco’’—UNLP,
CONICET—47 No 257 (B1900AJK) La Plata, Buenos Aires, Argentina
Received 10 October 2004; received in revised form 24 March 2005; accepted 30 March 2005
Available online 25 May 2005
Abstract
Two MnO2 were prepared by different methods: (i) MnCO3 decomposition; and (ii) oxidation of acid MnSO4 solution. The combined use
of X-ray diffraction (XRD), DRIFTS, DRS–UV–vis and TGA techniques revealed: (i) the presence of Mn4+–Mn3+; (ii) a poor crystallinity;
and (iii) OH species at temperatures higher than 200 8C.
The catalysts were tested in the complete oxidation of ethanol to carbon dioxide and water, and the reaction results showed that the
catalytic activity of both samples were larger than MnO2 (pyrolusite, b-MnO2) and Mn2O3 (like bixbyite) oxides. The high activity of the
prepared oxides can be related to the simultaneous presence of Mn3+ and Mn4+ ions and OH� species generated by Mn4+ vacancies. In
addition, in situ DRIFT spectroscopy demonstrated a different mechanism of adsorption-reaction of ethanol between the two synthesized
manganese oxides.
# 2005 Elsevier B.V. All rights reserved.
Keywords: g-MnO2 nsutite; Manganese oxides; DRIFTS; Ethanol oxidation; VOCs
www.elsevier.com/locate/apcatb
Applied Catalysis B: Environmental 61 (2005) 114–119
1. Introduction
Among the transition metal dioxides, MnO2 probably
exhibits the largest number of polymorph structures.
Pyrolusite (b-MnO2) is the most stable polymorph of
manganese dioxide and has the structure of rutile [1].
Ramsdellite (R-MnO2) is closely related to rutile except for
the fact that double chains replace the single chains of
edge—sharing octahedra [2] and nsutite or g-MnO2 has a
highly disordered structure and has been described as an
intergrowth of elements of pyrolusite and ramsdelllite [3].
Manganese dioxides can be classified according to the
number of MnO6 units and the number of MnO6 octahedral
chains between two basal layers to form tunnel openings.
They are usually symbolized T(m, n) where n and m stand
for the dimension of the tunnels in the two directions
perpendicular to the chains of edge-sharing octahedra [4].
Thus, pyrolusite is T(1, 1) and ramsdellite is T(1, 2), so that
* Corresponding author. Fax: +54 221 425 4277.
E-mail address: [email protected] (J.E. Sambeth).
0926-3373/$ – see front matter # 2005 Elsevier B.V. All rights reserved.
doi:10.1016/j.apcatb.2005.03.014
g-MnO2 could be represented as T(1, 1)–T(1, 2) inter-
growths.
Among the various polymorphs of MnO2 reported in the
literature (Chabre and Pannetier [4] have reported about 14
modifications) the g-MnO2 (nsutite) is the best-known
MnO2 polymorph used by the battery industry. g-MnO2 has
been extensively investigated for its important application as
cathode material for dry cells. According to McLean et al.
[5], structural and chemical defects are responsible for
electrochemical properties by the H+ insertion and increas-
ing the Fermi Level energy. The Volkenshtein Electronic
Theory [6] states that the catalytic activity is a function of
the electronic properties of solids, so the g-MnO2 phase
could be an excellent catalyst.
In this sense, Lahousse et al. [7] and Peluso et al. [8] have
shown that the g-MnO2 and one sample of Mn3+/g-MnO2,
respectively, are very suitable for VOC’s removal catalyst.
The aims of this paper are: (1) to characterize g-MnO2
catalysts prepared by different methods of synthesis; (2) to
study the catalytic activity in the complete oxidation of
ethanol to carbon dioxide and water, and to correlate this
L. Lamaita et al. / Applied Catalysis B: Environmental 61 (2005) 114–119 115
activity to the structural properties of the catalysts; and (3) to
propose a possible adsorption–oxidation mechanism.
2. Experimental
2.1. Sample preparation
Four manganese oxide samples were used in this work.
Two of them were prepared in order to obtain the g-MnO2
phase. In a first attempt to obtain this phase, MnCO3 (Riedel
de Haen RG) was oxidized at 500 8C under flowing oxygen
(50 cm3/min) with a heat rate of 20 8C min�1 [8]. This oxide
is named hereinafter M1.
The second manganese oxide (M2) used in this work was
prepared by the Netto et al. procedure [9]. A solution of
MnSO4�H2O dissolved in 3 M H2SO4 was oxidized with
flowing oxygen (30 cm3/min). pH was raised up to 3 with
NaOH. After being filtered and washed with deionised
water, the oxide was dried at 100 8C for 24 h.
For the purpose of comparison, a commercial reagent
grade MnO2 (Baker 99.9%), named M3, has been
investigated. Finally, a synthetic pyrolusite (M4) prepared
by thermal decomposition of solid Mn(NO3)2�4H2O was
also studied [10].
The list of the oxides, together with the relative notation,
is presented in Table 1.
2.2. Characterization
The structure of the studied samples was characterized
through X-ray diffraction (XRD) by using a Phillips PW
1390 instrument with Ni-filtered Cu Ka radiation
(l = 1.540589 A). The diffraction patterns were taken at
room temperature in the range 5 < 2u < 708. For phase
identification purposes, the JCPDS database of reference
compounds was used [11].
The BET specific surface areas were measured by N2
adsorption at the liquid nitrogen temperature (77 K) in a
Micromeritics Accussorb 2100 D sorptometer.
DRIFTS spectra were obtained in a Bruker IFS66
infrared spectrometer with a KBr optics and DTGS detector
in the 600–4000 cm�1 range. An environmental DRIFTS
chamber equipped with KBr window (Spectra Tech 0030-
103), allowing in situ treatments up to 300 8C and 1 atm was
coupled to the spectrometer. The spectra were obtained by
co-adding 200 scans collected at 4 cm�1.
Table 1
BET specific surface areas of the manganese oxides prepared in this study
Sample MnOx precursor BET specific area (m2 g�1)
M1 MnCO3 11
M2 MnSO4�H2O 55
M3a – 0.1
M4 Mn(NO3)2�4H2O 6.1a MnO2 baker 99%.
The DRS–UV–vis spectra were recorded on a Cary 2300
spectrophotometer in the region 200–2000 nm.
Thermal analyses (TG) of the catalysts were performed
on a DTA-TG 50 Shimadzu analyzer between room
temperature and 900 8C; small portions (14–16 mg) of the
samples were used at 10 8C min�1 in a flowing atmosphere
of air (30 cm3 min�1).
2.3. Catalytic activity
Ethanol was chosen as a probe molecule because it is an
important VOC produced in fermentation and sponge dough
processes and the industrial production of beer, food or
pharmaceutical products [12].
Catalytic test was carried out at atmospheric pressure in a
continuous flow tubular glass reactor. Two hundred
milligrams of the sample were loaded in form of fine
powder. The total gas flow of ethanol–air was 20 cm3 min�1
and the work temperatures were between 100 and 300 8C.
The feed composition was 900 ppm of ethanol in air, with an
excess of O2.
The reactants and the reaction products of a possible
incomplete reaction were analyzed by using an on-line gas
chromatograph (Shimadzu GC 8A, Porapak T for the
analysis of ethanol, CO2, acetaldehyde and a 5A zeolites for
the analysis of CO) and quantified by using a TCD. For the
purpose of comparison, the temperature at which the
conversion of the organic compound reaches 50% (T50) was
used to analyze the activity of the catalysts in VOCs
oxidation. The low T50 shows the best catalytic activity.
The adsorbed species formed during the reaction on the
manganese oxides were studied by in situ DRIFT spectro-
scopy by using the same experimental conditions achieved
in the studies of catalytic activity.
3. Results and discussion
X-ray diffraction was used to determine the bulk
crystalline phases in the samples. The diffraction patterns
of the manganese oxides are shown in Fig. 1. The XRD
Fig. 1. X-ray powder diffraction patterns of the manganese oxides.
L. Lamaita et al. / Applied Catalysis B: Environmental 61 (2005) 114–119116
Table 3
IR frequency vibration of manganese oxides
Sample IR bands (cm�1) References
M1 727 633 544
M2 718 633 542
M3 – 616 560
M4 – 615 560
b-MnO2 618 562 [19,20]
R-MnO2 743 649, 630 522 [19,20]
R-MnO2 740 687 529 [19]
g-MnO2 705 545 [19]
patterns of the commercial MnO2 sample (M3) show the
typical spectrum of b-MnO2 phase (pyrolusite) (JCPDS, 24-
0735). The oxide prepared by decomposition of
Mn(NO3)2�4H2O (M4) also shows the spectrum of
pyrolusite.
A more complex XRD pattern is obtained for M1 sample.
The signal at 2u 338 indicates the presence of a-Mn2O3,
bixbyite phase. Additionally, the broad and poorly resolved
peak in the 2u range 37–398 could indicate the presence of a
mixture of Mn(IV) and Mn(III) oxides, according to Arena
et al. [13]. Thus, this oxide is composed by a mixture of
Mn2O3 and an amorphous Mn4+–Mn3+ oxide.
Chabre and Pannetier [4] have reported that the XRD
spectrum of the g-MnO2 (nsutite) phase is always of rather
poor quality and consists, at best, of a small number of sharp
and broad lines on top of a diffuse background. The X-ray
powder diffraction pattern of M2 shows small and broad
lines at 2u 228, 378, 388 and 568, which reveals the presence
of the g-MnO2, according to database of the reference
compound (JCPDS, 17-0510).
The BET specific surface areas of the samples are
summarized in Table 1. The results showed a highest surface
area for M2 sample (nsutite phase oxide). Its high surface
area could be explained by a higher number of point defects
on this oxide surface [14]. As it can be seen from the table,
the M1 sample surface area is higher than that of M3 and
M4, but lower than M2. Finally, although M4 and M3
present the pyrolusite structure, M4 has a 60 times higher
surface area than M3.
The DRS–UV–vis position bands of different manganese
oxides can be seen in Table 2. For the M3 and M4 samples,
the spectra are characterized by an absorption band near
340 nm, which is assigned to Mn4+ [15]. The M1 and M2
samples also present that absorption. On the other hand, the
spectrum of M1 and M2 samples presents an absorption
band between 450 and 550 nm, which could be due to charge
transfer transitions of octahedral Mn3+ [16]. These features
suggest the existence of Mn4+ and Mn3+ in the bulk of the
M1 and M2 oxides, in agreement with XRD data.
The results of DRIFT spectroscopy of the studied
manganese oxides are reported in Table 3. Neither CO32�
nor SO42� groups were detected by DRIFTS spectroscopy in
the fresh catalysts.
IR spectroscopy has proven to be useful for the
identification of the nsutite phases [17], and yields more
reliable information than the X-ray diffraction technique. The
IR bands in the region 1000–400 cm�1 reveal information
Table 2
DRS–UV–vis bands of manganese oxides
Sample Mn4+ (nm) Mn3+ (nm)
M1 450–550 340
M2 450–550 340
M3 450–550
M4 450–550
Mn2O3 340
about MnO6 octahedra [18]. Moreover, Potter and Rosmann
[19] stated that the infrared spectroscopy is sensitive to
octahedral polymerization. Besides, Julien et al. [20]
observed a general decrease in band wavenumber with
increasing octahedral polymerization. These author assigned
the bands at 545 cm�1 in pyrolusite and at 522 cm�1 in
ramsdellite to the vibration due to the displacement of the
oxygen anions relative to the manganese ions along the
direction of the octahedral chains.
The M1 bands at 727 and 544 cm�1 could be assigned to
g-MnO2 and the band at 633 cm�1 to Mn2O3, respectively.
In the case of M2, DRIFTS bands are in accordance with
those reported by Potter and Rossman [19] for the g-MnO2
phase. In addition, these authors have also reported that the
characteristic bands of ramsdellite are at 522, 630, 649 and
743 cm�1. Taking into account the correlation of infrared
spectra made by Potter and Rossman [19] and Julien et al.
[20], it is possible to determine the structural characteristics
of our manganese oxides (M1, M2). In Fig. 2, the positions
of the IR bands are shown as a function of edge shared
octahedra. Both M1 and M2 samples are placed between the
ramsdellite and the pyrolusite, in agreement with the fact
that the g-MnO2 is a mixture of ramsdellite and pyrolusite
structures [3].
Fig. 2. Frequency positions of the IR bands, a function of the average MnO6
octahedral polymerization in manganese oxides. (ramsdellite bands, Potter
and Rosmann [18]).
L. Lamaita et al. / Applied Catalysis B: Environmental 61 (2005) 114–119 117
According to the results of the characterization techni-
ques we can say that:
(a) M
Fi
1 catalysts are constituted by an amorphous mixed
Mn2O3–g-MnO2 oxide.
(b) M
2 is an oxide of the g-MnO2 (nsutite) type withmanganese(III) and manganese(IV) cations.
(c) M
3 and M4 catalysts have the b-MnO2 (pyrolusite)structure and the manganese oxidation state is IV in both
catalysts.
Fig. 4. Total ethanol conversion as function of temperature.
Kanungo et al. [21] have reported that the thermal
analysis of manganese oxides between 25 and 900 8C in air
presents three different thermal effects: (a) loss of physically
adsorbed molecular water in the temperature range
100–200 8C; (b) loss of chemically bound water and a
change of crystalline phase between 200 and 400 8C; and (c)
the transformation of MnO2 to Mn2O3 around 550 8C.
Likewise, Vileno et al. [22] in their studies of OMS
manganese oxides also reported three thermal events: loss of
physisorbed water from 100 to 130 8C; loss of chemisorbed
water, inside or outside the tunnels, at 200 8C; and loss of
oxygen.
Fig. 3 depicts TG curves for the studied oxides except for
M4. As it can be seen, decomposition of MnO2 to Mn2O3 in
M1 sample occurs at 580 8C. The other catalysts show a
completely different behaviour. In fact, M2 sample is still
losing water even at 800 8C. This is in agreement with Petit
et al. [23] who pointed out that the g-MnO2 is non-
stoichiometric unless the temperature is raised up to 800 8C.
This is due to the presence of OH-groups in the bulk of the
oxide. They have also shown that this oxide presents
electrochemical activity at temperatures higher than 450 8C.
Finally, M3 is the most stable oxide. There is not physically
or chemically bound water, since no weight loss is present
until the temperature reaches 610 8C. At this stage, the
transformation to Mn2O3 occurs with approximately 10%
weight loss, and it is in agreement with the calculated value
(9.2%).
g. 3. TGA curves for manganese oxides: (a) M1; (b) M2; (c) M3.
The catalytic performances for ethanol oxidative
destruction on manganese oxide catalysts are shown as
function of temperature in Fig. 4. In all the samples tested,
CO2, H2O and acetaldehyde were formed during ethanol
oxidation, and no other partial oxidation products were
detected. Preliminary experiments showed that the com-
mercial manganese carbonate did not exhibit any activity at
all for the catalytic oxidation of ethanol.
Fig. 4 shows that the oxide with the nsutite structure (M2
sample) is a more active catalyst with a T50 more than 40 8Clower than that of the catalyst prepared by decomposition of
MnCO3 (M1 sample). Total combustion occurs at 200 8C for
the M2 catalyst, and at around 215 8C for M1 catalysts. In
both catalysts (M1 and M2), the selectivity to acetaldehyde
was at a peak when the conversion was low (less than 20%).
It was also observed that carbon dioxide was the only
product, when the conversion reached 30%. Both pyrolusite
type catalysts (M3 and M4) have the lowest activity of the
catalysts with a T50 close to 250 8C. The combustion is not
total until the temperature reaches 300 8C.
As we have seen, the most crystalline and lowest specific
surface area oxide (M3, b-MnO2) has the lowest activity,
and the most active catalyst (M2) has the lowest crystallinity
and the highest specific surface area. These results are in
agreement with Stobbe et al. [24] and Shaheen and Selim
[25], who have reported that the oxides with high degree of
crystallinity had a pronounced decrease in catalytic activity.
Nevertheless, not only the surface area but also the
structure of the oxides influences in the catalytic activity of
this series of oxides. M4 sample has a 60 times higher
specific surface area than that of M3 sample (Table 1), but
the activity of the former is lower than that of the latter (see
Fig. 4).
Besides the specific surface area and the crystallinity of
the catalysts, the existence of the Mn3+–Mn4+ couple [26–
28], the presence of cationic vacancies and OH groups [29]
are responsible for the high activity achieved by the nsutite-
phase oxide (M2).
L. Lamaita et al. / Applied Catalysis B: Environmental 61 (2005) 114–119118
As far as the oxidation state is concerned, the UV–vis and
DRIFTS spectra showed that M1 and M2 samples contained
Mn3+ and Mn4+ ions, whereas pure MnO2 without any
amount of Mn3+ constituted M3 and M4 samples. Although
both M1 and M2 samples contain Mn3+ in their structure,
they have different crystalline structures. In M1 sample,
Mn3+ is present in the form of crystalline Mn2O3 and it is
also present in a mixed Mn4+–Mn3+ amorphous oxide, where
the presence of the Mn4+ ions could be associated with g-
MnO2 phase. But in the case of M2 sample, Mn3+ cations
could be distributed in a nsutite matrix replacing Mn4+
cations. In this oxide, no Mn2O3 is detected.
Taking into account the catalytic activity results, oxides
with Mn valence between 3 and 4 were more active than
those with pure Mn (+4). M1 catalyst had a mixed structure
of Mn2O3 and g-MnO2, as it was confirmed by DRIFTS
spectra. Therefore, the presence of the couple Mn3+–Mn4+
could be responsible for the best catalytic performances
compared to b MnO2 and synthetic Mn2O3 [8].
The presence of Mn4+ vacancies in the nsutite structure is
followed by the formation of OH groups, in order to
equilibrate charges. The TGA results showed OH species at
200 8C when conversion was complete for M2. So, we can
say that OH groups played an important role in the oxidation
mechanism of VOC’s.
Fig. 5 shows the DRIFTS spectra of adsorbed species over
M1 catalysts. The spectrum of ethanol adsorption over M1
sample from RT to 70 8C shows bands assigned to ethoxide
species (1640–1650 cm�1 enolic anion; 1420 cm�1 dOH; near
1110 cm�1 C–O stretching and 1050 cm�1 ethoxy group)
[30,31]. At 70 8C, a new band at 1746 cm�1 attributed to
vibration nC O (acetaldehyde) is detected. At 150 8C, bands
begin to be formed at 1430 and 1555 cm�1, that can be
assigned to a chelate bidentate acetate [29]. Above 150 8C,
three bands are detected, two of them (1425–1570 and
Fig. 5. In situ DRIFTS spectra of the adsorption of ethanol on M1 catalyst
as a function of the temperature: (a) RT; (b) 70 8C; (c) 100 8C; (d) 120 8C;
(e) 150 8C; (f) 200 8C.
1045 cm�1) assigned to bridged bidentate acetate and at
1350 cm�1, a third band attributed to the monodentate acetate
species. The fact that different organic species are found in
M1 enables us to suppose that there are two different
mechanisms of adsorption-reaction, in good agreement with
Batiot and Hodnett [32] and Blasim-Aube et al. [33], who
have postulated that the ethanol oxidation to carbon dioxide
over catalyst can be described as:
Then, the formation of different adsorbed species
could depend upon the different bond dissociation enthalpy
of the weakest C–C and C–H bonds that ethanol and
acetaldehyde (389 kJ/mol C–H ethanol, 358 and 349 kJ/mol
C–C and C–H acetaldehyde) present when they are oxidized
to CO2.
Fig. 6 shows the infrared spectra of the adsorbed species
over M2 as a function of the reaction temperature. The
spectrum is different from the one recorded over M1 sample.
At room temperature, bands are detected at: (i) 1290 cm�1
attributed to adsorbed ethanol; (ii) 1023 cm�1 assigned to
ethoxy species; and (iii) 1475 cm�1 corresponding to a
vibration dCH2. In the range 70–200 8C, the main bands are
detected at: (i) 1270–1500 cm�1 attributed to monodentade
carbonate; and (ii) 1736 cm�1 assigned to carbonyl species
(acetaldehyde). In addition, the shoulder at 1177 cm�1 could
be due to ethoxy species coordinated over metal sites.
These results showed that the adsorption-reaction path-
way of ethanol over M2 was produced by the presence of
different adsorbed species with respect to M1 catalyst.
Taking into account these results, we propose the
following pathways for ethanol oxidation:
Fig. 6. In situ DRIFTS spectra of the adsorption of ethanol on M2 catalyst
as a function of temperature: (a) RT; (b) 70 8C; (c) 100 8C; (d) 150 8C; (e)
200 8C.
L. Lamaita et al. / Applied Catalysis B: Environmental 61 (2005) 114–119 119
The different mechanism could be attributed to the
formed solid. According to Ruetschi [34], g-MnO2 contains
Mn vacancies compensated by four protons that are attached
to the vacancy (OH formation). So, the interaction between
ethanol and oxygen and/or hydroxyl centres on the surface
of M2 catalyst could favour the formation of monodentade
carboxylate species and improve the catalytic performances.
4. Conclusions
Two manganese oxides were synthesized in order to
obtain the nsutite phase (g-MnO2) and to make a comparison
with two dioxides with the pyrolusite structure (b-MnO2).
The oxides with the nsutite (g-MnO2) structure (M2 and M1)
are more active catalyst than the solids with pyrolusite
structure (b-MnO2, M3 and M4).
The characterization results allow us to conclude that the
high activity of both catalysts could be due to its poor
crystallinity, the existence of the couple Mn3+–Mn4+ and the
presence of Mn4+ vacancies.
The in situ DRIFT spectroscopy showed that the two
synthetic manganese oxides presented different mechanisms
of adsorption-reaction of ethanol.
The mechanism of adsorption and the changes in the
activity between M1 and M2 catalysts could be explained
when assuming that these catalysts contain different
structures: Mn2O3/g-MnO2 and g-MnO2, respectively.
Acknowledgments
The authors acknowledge Prof. L. Cadus (UNSL,
Argentina) for TGA analysis, and Prof. G. Minnelli
(U. La Sapienza, Italy) for the DRS–UV–vis spectra. M.
Peluso thanks Comision de Investigaciones Cientıficas of
Buenos Aires (Argentina) for his Ph.D. fellowship.
References
[1] Baur, Acta Cryst. 32 (1976) 2200.
[2] T. Moore, M. Ellis, P. Selwood, J. Am. Chem. Soc. 72 (1950)
856.
[3] P. de Wolff, Acta Cryst. 12 (1959) 341.
[4] Y. Chabre, J. Pannetier, Prog. Solid State Chem. 23 (1995) 1.
[5] L. MacLean, Ch. Poinsignon, J. Amarilla, F. Le Cars, P. Strobel, J.
Mater. Chem. 5 (1995) 1183.
[6] F. Volkenshtein, The Electronic Theory of Catalysis on
Semiconductors, Pergamon Press, New York, 1963.
[7] C. Lahousse, A. Bernier, E. Gaigneaux, P. Ruiz, P. Grange, B. Delmon,
Third World Congress on Oxidation Catal, 1997, p. 777.
[8] M. Peluso, J. Sambeth, H. Thomas, React. Kinet. Catal. Lett. 80 (2003)
241.
[9] S. Netto, R. Hypolito, J. Valarelli, R. Giovanoli, R. Shultz-Gutler, An.
Acad. Bras. Cienc. 70 (1998) 563.
[10] J. Fernandez, B. Desai, K. Dalal, J. Power Sources 15 (1985)
209.
[11] Join Committee on Powder Diffraction Standard, JCPDS Files, Inter-
national Centre of Diffraction Data, 2000.
[12] A. Buonicore, W. Davids (Eds.), Air Pollution Engineering Manual,
Air and Waste Management Association, 1992.
[13] F. Arena, T. Torre, C. Raimondo, A. Parmaliana, Phys. Chem. Chem.
Phys. 3 (2001) 1911.
[14] J. Boyero Macstre, E.F. Lopez, J. Gallardo Amores, R. Ruano Casero,
V. Sanchez Escribano, E. Perez Bernal, Inorg. Mater. 3 (2001)
889.
[15] F. Millela, J. Gallardo Amores, M. Baldi, G. Busca, J. Mater. Chem. 8
(1998) 2525.
[16] A. Lahousse, P. Bernier, P. Grange, B. Delmon, P. Papaefthimou, T.
Ioannides, X. Verykiost, J. Catal. 178 (1998) 214.
[17] C. Julien, M. Massot, J. Power Sources 119–121 (2003) 743.
[18] M. Ananth, S. Pethkar, K. Dakshinamurthi, J. Power Sources 75
(1998) 278.
[19] R. Potter, G. Rosmann, Am. Mineral. 64 (1979) 1199.
[20] C. Julien, M. Massot, C. Poinsignon, Spectroch. Acta A 60 (2004)
689.
[21] S. Kanungo, J. Catal. 58 (1979) 419.
[22] E. Vileno, Y. Ma, H. Zhou, S. Suib, Microporous Mesoporous Mater.
20 (1998) 3.
[23] F. Petit, J. Durr, M. Lenglet, B. Hannoyer, Mater. Res. Bull. 28 (1993)
959.
[24] E. Stobbe, B. de Boer, J. Geus, Catal. Today 47 (1999) 161.
[25] W. Shaheen, M. Selim, Thermochim. Acta 332 (1998) 117.
[26] K. Parida, S. Kanungo, Thermochim. Acta 64 (1983) 131.
[27] N. Deraz, M. El-Sayed, A. El-Alal, Adsorption Sci. Technol. 19 (2001)
541.
[28] M. Galvan, P. O’Shea, J. Fierro, P. Arias, Catal. Commun. 4 (2003)
223.
[29] P. Ruetschi, R. Giovanoli, J. Electrochem. Soc. 135 (1988) 2663.
[30] W. Chun, C. Chuang, J. Lin, J. Phys. Chem. B 104 (2000)
8719.
[31] K. Nakamoto, Infrared and Raman Spectra of Inorganic and Coordi-
nation Compounds, fifth ed., Wiley & Son Inc., New York, 1997.
[32] C. Batiot, B. Hodnett, Appl. Catal. A 137 (1996) 179.
[33] V. Blasim-Aube, J. Belkouch, A. Monceaux, Appl. Catal. B 43 (2003)
175.
[34] P. Ruetschi, J. Electrochem. Soc. 131 (1984) 2737.